I want to solve this equation. please help me

1 回表示 (過去 30 日間)
kumar sourabh
kumar sourabh 2018 年 10 月 9 日
コメント済み: Alex Sha 2019 年 12 月 20 日
3.8 = a (1 - exp(b×8.2)) + c (exp(d×8.1) - 1);
27.9 = a (1 exp(b×63)) + c (exp(d×35.667) - 1);
23.5 = a (1 - exp(b×59.5)) + c (exp(d×59.5) - 1);
13.7 = a (1 - exp(b×35.7)) + c (exp(d×35.7) - 1)};

回答 (2 件)

madhan ravi
madhan ravi 2018 年 10 月 9 日
syms a b c d
eq1 = 3.8 == a*(1 - exp(b*8.2)) + c*(exp(d*8.1) - 1);
eq2 = 27.9 == a*(1 - exp(b*63)) + c*(exp(d*35.667) - 1);
eq3 = 23.5 == a*(1 - exp(b*59.5)) + c*(exp(d*59.5) - 1);
eq4 = 13.7 == a*(1 - exp(b*35.7)) + c*(exp(d*35.7) - 1);
[a,b,c,d]=vpasolve(eq1,eq2,eq3,eq4,a,b,c,d)
  10 件のコメント
Walter Roberson
Walter Roberson 2018 年 10 月 10 日
Looking again, there are a lot more points with very close residue than I had thought I had. Residues that are approximately 0.44011 were found for over 550 points, with a values ranging from 1337.8 to 171822.3, with b from -0.000388 to -2.965e-06, with c from 52.7 to 114.7, and d from -0.00244 to -0.00104 . a and b are practically linear with each other, and c and d are practically linear with each other, but the other pairs are only mostly linear with each other.
The residue was being calculated as
F = (a*(exp((41*b)/5) - 1) - c*(exp((81*d)/10) - 1) + 19/5)^2 + (a*(exp((119*b)/2) - 1) - c*(exp((119*d)/2) - 1) + 47/2)^2 + (a*(exp((357*b)/10) - 1) - c*(exp((357*d)/10) - 1) + 137/10)^2 + (a*(exp(63*b) - 1) - c*(exp((35667*d)/1000) - 1) + 279/10)^2
My tests did not show any sign of a "bowl" -- no minima that the values then rise again on the other side. I think doing better would require higher precision.
The marginally best combination I found was near 33762.2195025919 -1.50991408545558e-05 110.079764424126 -0.00109033033598055 but near 740.322188783004 -0.000712129376445178 37.1451062324705 -0.00368725773398294 is not far off
madhan ravi
madhan ravi 2018 年 10 月 11 日
Thank you once again @sir Walter.

サインインしてコメントする。


Alex Sha
Alex Sha 2019 年 12 月 19 日
Here is the right solution, obtained from 1stOpt:
a: -2.38880243947768E-9
b: 0.357245116257183
c: -32.3471147348507
d: -0.0154282661155712
  4 件のコメント
Walter Roberson
Walter Roberson 2019 年 12 月 19 日
The residue
F = (a*(exp((41*b)/5) - 1) - c*(exp((81*d)/10) - 1) + 19/5)^2 + (a*(exp((119*b)/2) - 1) - c*(exp((119*d)/2) - 1) + 47/2)^2 + (a*(exp((357*b)/10) - 1) - c*(exp((357*d)/10) - 1) + 137/10)^2 + (a*(exp(63*b) - 1) - c*(exp((35667*d)/1000) - 1) + 279/10)^2
is quite small, on the order of 1e-23 which is quite good for this situation.
Alex Sha
Alex Sha 2019 年 12 月 20 日
Hi, for "(a*(exp((41*b)/5) - 1) - c*(exp((81*d)/10) - 1) + 19/5)^2 + (a*(exp((119*b)/2) - 1) - c*(exp((119*d)/2) - 1) + 47/2)^2 + (a*(exp((357*b)/10) - 1) - c*(exp((357*d)/10) - 1) + 137/10)^2 + (a*(exp(63*b) - 1) - c*(exp((35667*d)/1000) - 1) + 279/10)^2", the minmum value:
Objective Function (Min.): 1.67130234831156E-23
a: -2.38880243946672E-9
b: 0.357245116257267
c: -32.347114734847
d: -0.015428266115563

サインインしてコメントする。

カテゴリ

Help Center および File ExchangeLinear Algebra についてさらに検索

タグ

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by